When a text-to-speech enumerator is created, it loads the registry entries for each engine installed on the system and creates an engine enumerator object to represent each engine.
The application calls the ITTSFind::Find member function on the text-to-speech enumerator with the address of a TTSMODEINFO structure that indicates the desired characteristics of the mode and the address of a TTSMODEINFORANK structure that indicates the relative importance of each characteristic. As an alternative, the application can call ITTSEnum member functions to enumerate the modes.
The text-to-speech enumerator passes these calls to each engine enumerator until one of them returns a mode that fits the requested characteristics.
After the application has created an engine object, the text-to-speech enumerator is no longer needed and the application can release it.